Introduction: There is no single blood marker for predicting the prognosis in ischemic stroke. A combination of multiple blood markers may enhance the ability to predict long-term outcome following ischemic stroke. Methods: Blood concentrations of neuronal markers (neuron-specific enolase, visinin-like protein 1, heart type fatty acid binding protein (hFABP) and neuroglobin), astroglial markers (S100B and glial fibrillary acidic protein), inflammatory markers (IL-6, TNF-α, and C-reactive protein), blood-brain barrier marker (matrix metalloproteinase 9), and haemostatic markers (D-dimer and PAI-1) were measured within 24 hours after stroke onset. The discrimination and reclassification for favorable and poor outcome were compared after adding individual or a combination of blood markers to the clinical model of stroke outcome. Results: In multivariate analysis, natural log-transformed (log) IL-6 (odds ratio (OR): 1.75, 95% CI: 1.25 to 2.25, P = 0.001) and loghFABP (OR: 3.23, 95% CI: 1.44 to 7.27, P = 0.005) were independently associated with poor outcome. The addition of a single blood marker to the clinical model did not improve the discriminating ability of the clinical model of stroke outcome. However, the addition of the combination of logIL-6 and loghFABP to the clinical model showed improved discrimination (area under receiver operating characteristic (AUROC) curve: 0.939 versus 0.910, P = 0.03) and reclassification performance (net reclassification improvement index: 0.18, P = 0.005). Conclusions: A combination of circulating IL-6 and hFABP level has an additive clinical value for the prediction of stroke outcome.